Abstract
Cuboidal [Fe4S4] clusters are ubiquitous cofactors in biological redox chemistry. In the [Fe4S4]1+ state, pairwise spin coupling gives rise to six arrangements of the Fe valences (‘valence isomers’) amongst the four Fe centers. How a protein active site dictates the arrangement of the valences in the ground state, as well as the population of excited-state valence isomers, is poorly understood in part because of the magnetic complexity of these systems. Here, we show that the ground-state valence isomer landscape can be simplified from a six-level system in an asymmetric protein environment to a two-level system by studying the problem in synthetic clusters [Fe4S4]1+ clusters with solution C3v symmetry. This simplification allows for the small energy differences between valence isomers (sometimes < 0.1 kcal/mol) to be quantified by simultaneously fitting the VT NMR and solution magnetic moment data. Using this fitting protocol, we map the excited state landscape for a range of clusters of the form [(SIMes)3Fe4S4–X/L]n, (SIMes = 1,3-dimesityl-imidazol-4,5-dihydro-2-ylidene; n = 0 for anionic, X-type ligands, and n = +1 for neutral, L-type ligands) and find that a single ligand substitution can alter the relative energies of valence isomers by at least 103 cm–1. On this basis, we suggest that one result of ‘non-canonical’ amino acid ligation in Fe–S proteins is to alter the distribution of the valence electrons in the manifold of thermally populated excited states.